Raman effect and IR absorption

Let's consider a molecule with N atoms, on which is shot a polychromatic IR beam. From the quantum selection rules, we know that in a single event of absorption only one of the 3N-6 oscillators can be excited or disexcited of one level. If we assume the oscillator in the ground state, then it absorbs from the radation a photon with a proper energy and goes in the second level.
I also have studied that in Raman effect, the visible radiation, which is more energetic than IR, excites the oscillator in a virtual state and then the oscillator relaxes down on the ground state or on another excited level by emetting a photon with energy corresponding to the jump.
So I don't understand why when an oscillator absorbs an IR photon then it doesn't relases it and comes back to its orginal state? I mean it is what happens during the Raman effect, why it doesn't happen also with IR? In this case I should not see any IR spectra since the same photon is absorbed and then emitted.

• Raman scattering happens in the IR as well, so I’m unclear on just what you are asking about? For instance, you can use stimulated Raman to generate a 1.8 micron coherent beam from Nd:YAG in hydrogen. – Jon Custer Dec 25 '18 at 22:35
• I'm just asking how is possbile to see IR spectra if, like it happens in Raman scattering, the oscillator after been excited relases a photon of the same energy when it comes back on the ground? I mean on the detector should arrive all the photons of the initial beam...@JonCuster – Landau Dec 26 '18 at 9:14

Maybe it will be helpful to refer to the Einstein $$A$$ and $$B$$ coefficients.
The bottom line is that absorption and stimulated emission do both occur. The appropriate rate coefficients ($$B$$) for the upward and downward processes are actually equal.
Spontaneous emission also occurs, at a rate governed by the $$A$$ coefficient. One can relate $$A$$ and $$B$$ by using a detailed balance argument; that's not critical to the present discussion, but it is explained on that Wikipedia page. The main point is that we only see a tiny fraction of those photons, the ones that happen to be emitted directly towards us, because the spontaneous emission occurs isotropically. So, even in an idealized picture where a steady state is maintained by equal rates of absorption and emission (spontaneous + stimulated) with no other relaxation processes, you would still expect to see an absorption line in the spectrum.